|Publication number||US4503556 A|
|Application number||US 06/364,930|
|Publication date||Mar 5, 1985|
|Filing date||Apr 2, 1982|
|Priority date||Apr 3, 1981|
|Also published as||CA1187184A, CA1187184A1, DE3113555A1, EP0067244A2, EP0067244A3|
|Publication number||06364930, 364930, US 4503556 A, US 4503556A, US-A-4503556, US4503556 A, US4503556A|
|Inventors||Wolfgang Scherl, Ludwig Abele, Friedrich Wahl|
|Original Assignee||Wolfgang Scherl, Ludwig Abele, Friedrich Wahl|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (3), Non-Patent Citations (4), Referenced by (78), Classifications (9), Legal Events (7)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The present invention relates to methods for automatic recognition of particular types of information appearing on a printed master, and in particular to a method for automatic recognition of white blocks as well as text, graphics and/or gray image areas on a printed master.
Automatic word processing units are increasingly being employed in office use for producing, modifying and storing written documents in an economic and time-saving manner. Such units have the capability of undertaking error corrections, insertion of new text passages, combining two or more texts having different origins, and random reproduction and electronic storage of the data corresponding to the text passages. The advantages of such automatic word processing units in comparison to conventional typewriters are the flexibility and time-saving in the production of written documents which can be generated by such units, and the higher efficiency resulting therefrom. A particularly time consuming step associated with the use of automatic word processing units is the transfer of information already existing on paper into the automatic word processing unit for storage and/or further processing. Manual transfer by keyboard of large amounts of text is extremely time consuming and accordingly various methods and devices have been developed for automatically transferring the information contained in texts into the word processing unit.
A problem in the automatic transfer of existing information contained in text passages into a word processing unit is that the master on which the text passages occur may also contain graphics and/or image areas. It is a problem in the art to automatically identify, classify and store these different types of information areas on a master in order to achieve an optimum coding of the data representing these different master areas as well as to permit separate manipulation of the data representing those areas within the word processor.
It is an object of the present invention to provide a method for automatically recognizing and classifying white blocks as well as text, graphics and/or gray image areas on a printed master in an economic and reliable manner without manual intervention.
The above object is inventively achieved in a method in which the printed master is optoelectronically scanned and the analog signals resulting from the scanning are converted into digital signals. The digital signals are employed for generating a binary image of the printed master. The binary image is searched for white areas and in a subsequent step contiguous printed master areas which are completely enclosed by white areas are identified and assigned a number for the purposes of subsequent identification. Each such area which has been located is surveyed for characterizing features which are then analyzed, such as by statistical methods, in order to classify the area in question as a white area, a text area, a graphics area or a gray image area. The entire printed master is analyzed in this manner with the result that the digital information representing the various areas of the master are respectively optimally encoded.
In one embodiment of the method, the binary image is searched for white blocks by generating separate binary images respectively containing horizontal white blocks and vertical white blocks and the two such binary images are subsequently linked by means of a logical function such that a composite binary image arises which shows the printed master areas which are surrounded by "white frames." The logical function may be, for example, an OR function.
The method disclosed herein provides a relatively simple and economic means for classifying types of information found on a printed master in a manner which requires relatively low memory and data processing outlay in which the data processing means may be, for example, a microcomputer.
FIG. 1 is a block diagram of an apparatus for executing the method disclosed herein.
FIG. 2 is a flow chart showing the individual method steps of the method disclosed herein.
FIG. 3 is a section of a typical printing master to which the method disclosed herein can be applied.
FIG. 4 is a reproduction of the typical printing master shown in FIG. 3 after execution of a preparatory method step.
FIG. 5 is a reproduction of the typical printing master shown in FIG. 3 after execution of a first method step.
FIG. 6 is a reproduction of the printing master shown in FIG. 3 after the execution of a second method step.
FIG. 7 is a reproduction of the printing master shown in FIG. 3 after the execution of a third method step.
FIG. 8 is a reproduction of the printing master shown in FIG. 3 after the execution of a fourth and last method step.
FIG. 9 is a graph showing the distribution of selected features of individual segmented master areas according to FIG. 7.
FIG. 1 is a block diagram of a preferred embodiment of an apparatus for executing the method disclosed herein. The apparatus includes a video scanner V for optoelectronically scanning a printed master D and for generating analog signals corresponding to the information contained on the master D. The master D may contain white areas, text areas, graphics areas and/or gray image areas.
The analog output signals of the video scanner V are supplied to an analog/digital converter A/D. The digital signal outputs of the converter A/D are supplied to the signal imput of a memory S which is preferably a semiconductor memory. A computer R, which may be a microcomputer or a microprocessor, has access to the memory S. The computer R processes the data deposited in the memory S according to the method disclosed herein as described in detail below in order to determine the characteristics of the various portions of the printed master D for classifying those areas as text, graphics or image areas. The entire process for undertaking such a classification is set forth in individually labelled steps in the diagram shown in FIG. 2, which includes some initial preparatory steps which are executed in a manner known to those skilled in the relevant technology. The first step, labelled a in FIG. 2, is that of optoelectronically scanning the master. Step b of FIG. 2 is the step of reading the signals generated by the scanning step a into the memory S. A gray scale value is assigned to the signals which have been read into the memory S in step c. The particular gray scale employed for assigning the gray scale values can be selected depending upon the particular type of printing masters to be evaluated. The information which may be expected on a typical master D is shown in FIG. 3 consisting of text, graphics, image and white areas.
In a final preparatory method step d, a binary image of the printed master is generated with each analog signal representing the optoelectronic imaging of a scanned point allocated a binary number in a decision operation with a first binary number, such as "1", representing a white level and a second binary number, such as "0", representing a black level. For the printed master shown in FIG. 3, a binary image as shown in FIG. 4 results after method step d.
In a first step e of the method disclosed and claimed herein, the binary image of the printed master of the type shown in FIG. 4 is searched for white areas. This may be undertaken in the horizontal direction of the binary image, in which case the partial image shown in FIG. 5 arises, as well as in the vertical direction of the binary image, in which case another partial binary image arises as is shown in FIG. 6.
In a second step f, the existence of contiguous areas of the printed master which are completely enclosed or framed by white regions is determined as is shown in FIG. 7 such as, for example, by combining FIGS. 5 and 6 by an exclusive OR logic operation. Such "white framed" areas are numbered for the purpose of subsequent identification in a further method step f', however, the numbers are assigned only in the memory S and therefore are not shown in FIG. 7. In a third method step g, each "white framed" area of the printed master is surveyed by known means and selected characterizing features of the area in question are extracted and are subsequently analyzed in a final method step h, such as by statistical analysis, for classifying the area in question as a white area, a text area, a graphics area, or a gray image area. The different types of areas into which the master may be segregated are represented in FIG. 8 by three different sized cross-hatchings, with white areas being indicated by the absence of cross-hatching. After such classification, an optimum encoding of the informational content of each of the respective printed areas is undertaken.
The signal generated in analog form representing the optoelectronic imaging of the information carried on the master D may be separated into white levels and black levels by known analog methods; however, in a preferred embodiment the classification is undertaken by first digitizing the analog signal and subsequently classifying the digital values as representing white levels or black levels by comparison with a predetermined digital threshold value. The subdivision of the scanning signals into white levels and black levels may also be undertaken by comparison with a so-called gray threshold by which the binary image of the type shown in FIG. 4 is generated. The gray threshold may be a value which is, for example, 80% of the maximum possible white level.
In a further embodiment of the method, a binary image with horizontal white blocks and a binary image with vertical white blocks are formed from the master and the two binary images are linked to one another by means of a logical function so that a combined binary image results which shows the existence and location of "white frames" which may be present and which enclose various areas of the printed master. The logic function linking the two separately-generated binary images is preferably an OR function.
A number representing a gray scale value is assigned to each of the printed master areas enclosed by a white frame which has been identified. This number identifies the respective area as a text area, a graphics area, a gray image area or a white block.
In a further embodiment of the method, a percentage of the white points in a particular area of the printed master and a percentage of pairs of points of the same binary value in the same printed master area are determined, each pair of points representing two adjacent points in the line direction. The features gained in this manner are employed as is known to those skilled in the art for analyzing the various areas of the printed master to classify the areas by the use of a so-called piece-by-piece linear classifier. The distribution of such features of a particular area of the printed master shown in FIG. 7 is graphically illustrated in FIG. 9 wherein a pair of features PW /PCC is represented per printed master area. In FIG. 9 the numerical percentage of white points is labelled PW and the numerical percentage of pairs of points with the same binary value is labelled PCC. Those points falling within the area labelled P are classified as a gray image area, those points falling within the area labelled T are classified as a text area, those points falling within the portion of the graph labelled G are classified as a graphics area, and those points falling within the area of the graph labelled W are classified as a white block.
In addition to the classification graphically represented in FIG. 9, an additional step h' may be undertaken in accordance with the method disclosed herein by which the spatial relationship of all of the areas classified on the basis of the analysis shown in FIG. 9 is further examined in comparison with typical spatial arrangements of such areas on typical printed masters. If a non-correlation with a typical arrangement is determined, an appropriate correction in step h' of the classification is undertaken to reclassify certain of the areas until correlation at a selected level is achieved.
After the various areas of the printed master have been classified in accordance with the method described above, different encoding methods can be applied to the different printed master areas which is preferably the least redundant type of encoding for the particular class of the informational content of the printed master area.
The method disclosed and claimed herein provides a relatively fast processing classification of the information carried on a printed master which employs relatively simple processing procedures which can be undertaken by state of the art computers and peripheral devices. The necessity of random memory access is eliminated permitting the use of economical memory and computer devices.
Although modifications and changes may be suggested by those skilled in the art it is the intention of the inventors to embody within the patent warranted hereon all such changes as reasonably and properly come within the scope of their contribution to the art.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US3496543 *||Jan 27, 1967||Feb 17, 1970||Singer General Precision||On-line read/copy data processing system accepting printed and graphic material|
|US3805237 *||Apr 30, 1971||Apr 16, 1974||Ibm||Technique for the conversion to digital form of interspersed symbolic and graphic data|
|US4411015 *||May 19, 1981||Oct 18, 1983||Siemens Aktiengesellschaft||Method and apparatus for automatic recognition of image and text/graphics areas on a master|
|1||"Procedures for an Automatic Segmentation of Text, Graphics and Half Tone Regions in Documents", by Oja et al.|
|2||*||Procedures for an Automatic Segmentation of Text, Graphics and Half Tone Regions in Documents , by Oja et al.|
|3||*||Proceedings of the Second Scandinavian Conference on Image Analysis, 6/17/81, pp. 177 182.|
|4||Proceedings of the Second Scandinavian Conference on Image Analysis, 6/17/81, pp. 177-182.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US4668995 *||Apr 12, 1985||May 26, 1987||International Business Machines Corporation||System for reproducing mixed images|
|US4750209 *||Dec 12, 1986||Jun 7, 1988||Ricoh Company, Ltd.||System for processing an image having both letter and photographic information|
|US4769849 *||Dec 19, 1985||Sep 6, 1988||The Palantir Corporation||Method and apparatus for separating overlapping patterns|
|US4899394 *||May 9, 1986||Feb 6, 1990||Prodigy Systems Corporation||Apparatus and method for image compression|
|US4941125 *||Aug 1, 1984||Jul 10, 1990||Smithsonian Institution||Information storage and retrieval system|
|US5018024 *||Nov 17, 1989||May 21, 1991||Canon Kabushiki Kaisha||Image processing apparatus using two-dimensional space frequency to discriminate image type|
|US5025481 *||Oct 4, 1989||Jun 18, 1991||Ricoh Company, Ltd.||Dot region discriminating method|
|US5038381 *||Mar 19, 1990||Aug 6, 1991||New Dest Corporation||Image/text filtering system and method|
|US5048096 *||Dec 1, 1989||Sep 10, 1991||Eastman Kodak Company||Bi-tonal image non-text matter removal with run length and connected component analysis|
|US5048109 *||Dec 8, 1989||Sep 10, 1991||Xerox Corporation||Detection of highlighted regions|
|US5065257 *||Sep 18, 1990||Nov 12, 1991||Canon Kabushiki Kaisha||Image processing apparatus with means for compensating for dust., etc., on an original|
|US5065437 *||Dec 8, 1989||Nov 12, 1991||Xerox Corporation||Identification and segmentation of finely textured and solid regions of binary images|
|US5073953 *||Sep 1, 1989||Dec 17, 1991||Oce Nederland B.V.||System and method for automatic document segmentation|
|US5131049 *||Dec 8, 1989||Jul 14, 1992||Xerox Corporation||Identification, characterization, and segmentation of halftone or stippled regions of binary images by growing a seed to a clipping mask|
|US5134572 *||Nov 18, 1988||Jul 28, 1992||Hitachi Software Engineering Co., Ltd.||Partitioned image processing system for an image processing device|
|US5140650 *||Feb 2, 1989||Aug 18, 1992||International Business Machines Corporation||Computer-implemented method for automatic extraction of data from printed forms|
|US5148495 *||May 15, 1991||Sep 15, 1992||Ricoh Company, Ltd.||Line region segmentation method|
|US5157740 *||Feb 7, 1991||Oct 20, 1992||Unisys Corporation||Method for background suppression in an image data processing system|
|US5193122 *||Dec 3, 1990||Mar 9, 1993||Xerox Corporation||High speed halftone detection technique|
|US5202933 *||Dec 8, 1989||Apr 13, 1993||Xerox Corporation||Segmentation of text and graphics|
|US5235654 *||Apr 30, 1992||Aug 10, 1993||International Business Machines Corporation||Advanced data capture architecture data processing system and method for scanned images of document forms|
|US5251273 *||Apr 15, 1992||Oct 5, 1993||International Business Machines Corporation||Data processing system and method for sequentially repairing character recognition errors for scanned images of document forms|
|US5267330 *||Dec 24, 1990||Nov 30, 1993||Canon Kabushiki Kaisha||Image processing apparatus|
|US5272764 *||Aug 28, 1991||Dec 21, 1993||Xerox Corporation||Detection of highlighted regions|
|US5278918 *||Nov 27, 1991||Jan 11, 1994||Caere Corporation||Optical character recognition method and apparatus using context analysis and a parsing algorithm which constructs a text data tree|
|US5278919 *||Feb 27, 1992||Jan 11, 1994||Nec Corporation||Image area determining apparatus|
|US5285504 *||Sep 27, 1991||Feb 8, 1994||Research Foundation Of The State University Of New York||Page segmentation with tilt compensation|
|US5305396 *||Apr 17, 1992||Apr 19, 1994||International Business Machines Corporation||Data processing system and method for selecting customized character recognition processes and coded data repair processes for scanned images of document forms|
|US5313526 *||Dec 30, 1991||May 17, 1994||Goldstar Co., Ltd.||Method for disconnecting character strings of a compressed image|
|US5325441 *||Jan 31, 1992||Jun 28, 1994||Westinghouse Electric Corporation||Method for automatically indexing the complexity of technical illustrations for prospective users|
|US5335290 *||Apr 6, 1992||Aug 2, 1994||Ricoh Corporation||Segmentation of text, picture and lines of a document image|
|US5369507 *||Aug 19, 1992||Nov 29, 1994||Konica Corporation||Image processing apparatus and method capable of discriminating between character/photograph areas by finding blank-stripe zones and dividing the areas into blocks|
|US5394487 *||Oct 27, 1993||Feb 28, 1995||International Business Machines Corporation||Forms recognition management system and method|
|US5410619 *||Jun 14, 1994||Apr 25, 1995||Ricoh Company, Ltd.||Digital image processing apparatus for processing a variety of types of input image data|
|US5428694 *||Oct 14, 1993||Jun 27, 1995||International Business Machines Corporation||Data processing system and method for forms definition, recognition and verification of scanned images of document forms|
|US5430808 *||Apr 17, 1992||Jul 4, 1995||At&T Corp.||Image segmenting apparatus and methods|
|US5467410 *||Mar 20, 1992||Nov 14, 1995||Xerox Corporation||Identification of a blank page in an image processing system|
|US5473444 *||Mar 27, 1995||Dec 5, 1995||Ricoh Company, Ltd.||Image processing device having function of detecting dotted-image area|
|US5506918 *||Dec 23, 1992||Apr 9, 1996||Kabushiki Kaisha Toshiba||Document skew detection/control system for printed document images containing a mixture of pure text lines and non-text portions|
|US5521991 *||Oct 29, 1993||May 28, 1996||International Business Machines Corporation||Method and system for fast forms recognition of document form images|
|US5548664 *||Jun 29, 1994||Aug 20, 1996||Wang Laboratories, Inc.||Automatic determination of blank pages and binary images' bounding boxes|
|US5555325 *||Oct 22, 1993||Sep 10, 1996||Lockheed Martin Federal Systems, Inc.||Data capture variable priority method and system for managing varying processing capacities|
|US5566255 *||May 23, 1995||Oct 15, 1996||Ricoh Company, Ltd.||Segmenting a page of a document into areas which are text and areas which are halftone|
|US5596655 *||Jun 7, 1995||Jan 21, 1997||Hewlett-Packard Company||Method for finding and classifying scanned information|
|US5613016 *||Dec 28, 1994||Mar 18, 1997||Ricoh Company, Ltd.||Area discrimination system for text image|
|US5617485 *||Apr 4, 1995||Apr 1, 1997||Ricoh Company, Ltd.||Image region segmentation system|
|US5619592 *||Jun 7, 1995||Apr 8, 1997||Xerox Corporation||Detection of highlighted regions|
|US5625721 *||Apr 6, 1994||Apr 29, 1997||Matsushita Information Technology Laboratory||Certifiable optical character recognition|
|US5696843 *||May 12, 1995||Dec 9, 1997||Sharp Kabushiki Kaisha||Automatic image quality controlling apparatus for use in an electronic copier|
|US5703972 *||Nov 17, 1995||Dec 30, 1997||Panasonic Technologies, Inc.||Certifiable optical character recognition|
|US5757957 *||Mar 1, 1995||May 26, 1998||Ricoh Company, Ltd.||Apparatus and method for area separation for image, having improved separation accuracy|
|US5818607 *||Feb 21, 1996||Oct 6, 1998||Hitachi, Ltd.||Image signal processing apparatus and information transmission/reception apparatus|
|US5832114 *||Jun 17, 1994||Nov 3, 1998||Canon Kabushiki Kaisha||Facsimile apparatus for receiving and transmitting standardized video images|
|US5832132 *||Oct 6, 1995||Nov 3, 1998||Canon Kabushiki Kaisha||Image processing using neural network|
|US5881168 *||Apr 7, 1995||Mar 9, 1999||Canon Kabushiki Kaisha||Image processing apparatus method and system for converting character codes of character images extracted from an original image and positions of the images for transmission|
|US5892843 *||Jan 21, 1997||Apr 6, 1999||Matsushita Electric Industrial Co., Ltd.||Title, caption and photo extraction from scanned document images|
|US6002798 *||Jan 19, 1993||Dec 14, 1999||Canon Kabushiki Kaisha||Method and apparatus for creating, indexing and viewing abstracted documents|
|US6005680 *||Apr 4, 1995||Dec 21, 1999||Canon Information Systems, Inc.||Method for capturing a document image, a scanner using the method and a document image management system using the scanner|
|US6009198 *||Nov 21, 1997||Dec 28, 1999||Xerox Corporation||Method for matching perceptual shape similarity layouts across multiple 2D objects|
|US6101274 *||Jun 2, 1997||Aug 8, 2000||Siemens Corporate Research, Inc.||Method and apparatus for detecting and interpreting textual captions in digital video signals|
|US6192152 *||May 21, 1993||Feb 20, 2001||Canon Kabushiki Kaisha||Image processing apparatus|
|US6208744 *||Dec 11, 1995||Mar 27, 2001||Casio Computer Co., Ltd.||Document image processor and method for setting a document format conforming to a document image|
|US6298173||Oct 3, 1997||Oct 2, 2001||Matsushita Electric Corporation Of America||Storage management system for document image database|
|US6411394||Oct 6, 1998||Jun 25, 2002||Hitachi, Ltd.||Image signal processing apparatus and information transmission/reception apparatus using the same|
|US6449065||May 7, 1999||Sep 10, 2002||Canon Kabushiki Kaisha||Method for capturing a document image, a scanner using the method and a document image management system using the scanner|
|US6832726||Dec 12, 2001||Dec 21, 2004||Zih Corp.||Barcode optical character recognition|
|US7305612 *||Mar 31, 2003||Dec 4, 2007||Siemens Corporate Research, Inc.||Systems and methods for automatic form segmentation for raster-based passive electronic documents|
|US7311256 *||Nov 30, 2004||Dec 25, 2007||Zih Corp.||Barcode optical character recognition|
|US7337399||Jan 3, 2005||Feb 26, 2008||Infoprint Solutions Company||Apparatus, system, and method for editing a region of a document intersecting multiple content component types in a single operation|
|US7660014||Jun 6, 2006||Feb 9, 2010||Konica Minolta Business Technologies, Inc.||Image processing apparatus capable of extracting rule from document image with high precision|
|US7676090 *||Jun 2, 2006||Mar 9, 2010||Lexmark International, Inc.||Systems and methods for content-based document image enhancement|
|US8208744 *||Jun 6, 2006||Jun 26, 2012||Konica Minolta Business Technologies, Inc.||Image processing apparatus capable of accurately and quickly determining character part included in image|
|US20030163785 *||Feb 28, 2002||Aug 28, 2003||Hui Chao||Composing unique document layout for document differentiation|
|US20040194035 *||Mar 31, 2003||Sep 30, 2004||Amit Chakraborty||Systems and methods for automatic form segmentation for raster-based passive electronic documents|
|US20050077354 *||Nov 30, 2004||Apr 14, 2005||Karl Torchalski||Barcode optical character recognition|
|US20060280366 *||Jun 2, 2006||Dec 14, 2006||Lexmark International, Inc.||Systems and Methods for Content-Based Document Image Enhancement|
|US20070165280 *||Jun 6, 2006||Jul 19, 2007||Konica Minolta Business Technologies, Inc.||Image processing apparatus capable of extracting rule from document image with high precision|
|US20070172137 *||Jun 6, 2006||Jul 26, 2007||Konica Minolta Business Technologies, Inc.||Image processing apparatus capable of accurately and quickly determining character part included in image|
|U.S. Classification||382/176, 358/462, 382/180|
|International Classification||G06T7/40, H04N1/04, G06K9/20, H04N1/40|
|Mar 14, 1985||AS||Assignment|
Owner name: SIEMENS AKTIENGESELLSCHAFT, BERLIN AND MUNICH, A G
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST.;ASSIGNORS:SCHERL, WOLFGANG;ABELE, LUDWIG;WAHL, FRIEDRICH;REEL/FRAME:004372/0803
Effective date: 19841227
|Oct 4, 1988||REMI||Maintenance fee reminder mailed|
|Nov 23, 1988||FPAY||Fee payment|
Year of fee payment: 4
|Nov 23, 1988||SULP||Surcharge for late payment|
|Oct 8, 1996||REMI||Maintenance fee reminder mailed|
|Mar 2, 1997||LAPS||Lapse for failure to pay maintenance fees|
|May 13, 1997||FP||Expired due to failure to pay maintenance fee|
Effective date: 19970305